shinyloadtest

50 simulated users executing back to back sessions. Warmup or cooldown sessions (that are desaturated) start before or end after the vertical dotted line. Note that narrower event bars mean better performance.

925 maintenance sessions ordered from fastest to slowest completion time. The red line marks how long the original recording session took to complete (~33s). Sessions should end around the same time as each other for consistent behavior.

Each session is represented with a line that cascades downward from each event. As time progresses, the line will move to the right. Consistent behavior will be represented with parallel lines, whereas inconsistent behavior will have lines jut arbitrarily to the right. Warmup and cooldown sessions are included as grey lines.

Total HTTP and supporting files (JS/CSS) latency (load time) is displayed for each session and run combination. Each vertical bar represents the amount of time a session's Shiny application had to wait before it is able to ask the server for information. A resonable amount of time to wait as been set to 5s. Warmup and cooldown sessions are displayed outside the dotted maintenance period lines.

Maximum WebSocket latency (calculation time) is displayed for each session and run combination. The maximum time is shown to convey the longest amount of time a user would have to wait for a response from the Shiny server. Warmup and cooldown sessions are displayed outside the dotted maintenance period lines.

Event plots are arranged by the slowest maximum time within each plot.

Event plots are arranged by the slowest minimum time within each plot.

Event plots are arranged by the largest slope magnitude found when fitting a line to each run.

Event plots are arranged by the largest intercept magnitude found when fitting a line to each run

Event plots are arranged by the largest error magnitude found when fitting a line to each run.